4 METHODOLOGYANDRESEARCHPROCESS
4.2 Studydesign
4.2.1 Sampling
42 As seen above, part of our own preconception is associated with our personal experience. This is in turn tinged by the theoretical and epistemological understanding that I as a researcher have of my field of study. Chapters 2 and 3 form part of this whole.
43 4.2.1.1 Nursing homes
Criteria for selection of nursing homes:
x Public nursing homes, including those that are part of the municipal health and care services in Southern Norway.
x Nursing homes in various geographical locations, in cities (>200 000), cities (<100 000), small built-up areas/small towns (>6000) and rural areas (<6000).
x Nursing homes with a minimum of one somatic unit and a minimum of one dementia unit.
Table 4.1: Nursing homes – demographic data
Variable: Location Number
of patients
Number of wards in the nursing home (dementia units)
Full-time equivalence - physicians`
(1=100%)
Full-time equivalence - nurses`
(1=100%) Nursing
home 1
city (< 100 000) 72 3 (1) 0.6 15
“ 2 city (> 200 000) 108 5(1) 0.8 16
“ 3 small town 22 2 (1) 0.2 11
“ 4 rural area 62 3(1) 1.0 17.13
“ 5 small town 36 4 (1) 0.2 8.5
“ 6 small town 31 4 (1) 0.3 13
” 7 city (< 100 000) 59 7 (2) 0.6 19.5
“ 8 city (< 100 000) 40 2 (1) 0.6 13.27
“ 9 city (< 100 000) 58 3 (1) 0.4 15.75
“ 10 city (> 200 000) 96 6 (2) 0.9 19
A total of ten nursing homes in various communities in Southern Norway were contacted in line with purposive sampling as the setting for the study. These ten had a desirable
geographical distribution. I took into consideration that certain cultural variations in terms of family traditions might occur, including the relationship to and experience of death according to the location of the nursing home. I also wanted to include nursing homes with varying distances to a hospital, since I assumed that this might have an effect on the propensity to hospitalise nursing-home patients. Furthermore, for nursing homes located in small rural
44 communities there is a greater probability that the children of elderly patients will live some distance away as a result of centralisation over the last centuries. I wanted to include this factor, because the geographical proximity of relatives to the nursing home may have an influence on the decision-making process.
The management of the health and care services in municipalities where relevant nursing homes were identified were contacted by telephone.
The table 4.1shows the characteristics of the nursing homes, in terms of geographical location, the number of patients and their staffing by doctors and nurses.
4.2.1.2 Relatives
Criteria for selection of next of kin:
x Relatives of patients older than 75 years
x Relatives of patients with fully or partly absent competence to provide consent at the final stage of life (See Pt. 4.3.2)
x Relatives of patients for whom life-prolonging treatment at the final stage of life had been considered
x Son, daughter, son-in-law, daughter-in-law, spouse/partner x Equal distribution of men and women
x Relatives who were resident in the same area as the patient (the nursing home) and relatives who were resident more than 100 km from the patient (the nursing home) x Relatives from a variety of educational backgrounds
x Norwegian-speaking relatives
x Requirement for the interview to take place no less than two months and no more than one year after the death of the patient.
The sample of relatives was contacted deliberately to identify information-rich informants (purposive sampling)(Patton 2002, Denzin and Lincoln 2005). My goal was also to achieve a gender balance in the sample of relatives, and this could be ensured within the framework of ten nursing homes. Men and women might give different information about the topic due to social differences in family relations in care for the elderly. Table 4.2 shows the relatives` demographic data, linked to the patients.
45 Table 4.2: Relatives - demographic data, linked to patient data
Location of the nursing home
Relative Patient
M/F * Age Relation to the informant
Age Dementia Y/N **
Time in the nursing home
Hospitalisation from the nursing home
City (<100 000) M 65 Father Mother
96 88
N Y
Short 3 years
Admission, apoplexy Admission, pneumonia City (<100 000) F 44 Mother 84 Y 3 years Admission, fall x-ray City (>200 000) F 62 Mother 92 Y 2.5 years Admission, apoplexy
Small town F 69 Mother 95 Y 1.5 years No
Small town F 62 Father
Mother 85 85
N Y
5 months Alive
No
Rural area M 41 Father 78 Y 6 years No
Rural area M 63 Father
Mother 83 80
N Y
10 days 3 years
No
Admission, acute abdominal pain
Small town M 60 Father 90 N 3 years Admission, apoplexy
Small town M 55 Mother
Father 88 90
N Y
1 year 2 years
Admission, pneumonia Admission, pneumonia City (<100 000) F 60
Mother-in-law Mother
94 92
-language N
10 days Short-term placement
No
Admission UVI (died in hospital shortly after)
City (<100 000) F 62 Mother 90 N -language
2 months Admission, apoplexy
City (<100 000) M 60 Mother 89 N 1.5 years No
City (<100 000) M 77 Wife 76 Y 5.5 years Admission, pneumonia
City (<100 000) F 43 Father 83 N 5 months No
City (>200 000) M 55 Mother 78 Y 3 years No
*M/F= Male/Female ** Y/N= Yes/ No
46 4.2.1.3 The sample of physicians and nurses
Criteria for the selection of physicians:
x At least two years of continuous employment as a physician in the nursing home Criteria for the selection of nurses:
x At least two years of continuous employment as a nurse in the nursing home x At least 50 per cent of a full-time equivalent position
x Clinical nurses on rota without management responsibility at the department level Selecting the ten nursing homes on the basis of geographical criteria left certain factors uncontrollable with regard to the physicians and nurses. A factor which appears to be randomly distributed, and which can be seen as beneficial in terms of purposive sampling, is the age distribution among the physicians and nurses. The same observation applies to their respective years of experience as health workers in nursing homes. In eight of the ten nursing homes, the physicians’ working hours were controlled by their General Practitioner (GP) contract with the local authorities, which required them to spend 20 per cent of their working week in the nursing home. Two of the physicians were employed in permanent positions at the nursing home, on a 60 per cent and 90 per cent basis respectively. This distribution reflects the staffing rate of physicians in Norwegian nursing homes (Statistics Norway SSB29). None of the physicians were specialists in geriatrics, eight of nine were GPs. The last was a surgeon.
All nursing jobs were 0.7 full-time equivalents or above. Among the nurses, two had
specialist training in palliation and one in geriatric nursing. Table 4.3 shows the demographic data of physicians and nurses.
Table 4.3: Physicians and nurses – demographic data
Variable Physicians Nurses
No. 9 10
Age (mean) 44 50.5
Experience (years) 11 9.8
(*= mean experience as an employee in a nursing home)
29 http://www.ssb.no/helsetilstand_en/
47